How to make great mixes

[Mostess]

I am practically paralyzed by the mixing possibilities of my new digital recording system. I grew up with a four track, so after all the recording was done, I had eight knobs to deal with (four levels, four pans). That's about the size of my mixing brain.

Now I have nearly as many tracks as I want. Each has a global level, pan, and EQ. Each can have as many effects as I want including additional EQ, dynamic panning, dynamic levels. Not to mention compression; that's a whole world of confusion right there; too much makes my ears hurt, but too little is so obviously not enough! Not to mention stuff I can barely understand: notably stereo expansion and phase shifting.

How does Josh Woodward make such clear mixes? How did Blue put together that mix of Octothorpe's children's song so that all the muddled insanity that is the brilliance of Octothorpe is clean and internally consistent? What can I do to turn the signature muffled, thin, clumsy Hostess Mostess sound into the wish-it-were-signature sharp, crystal clear, smooth, delicious Hostess Mostess sound?

Someone teach me the magic words!

[joshw]

How does Josh Woodward make such clear mixes?

By clearing out the junk. I used to use a bunch of crap to try and beef up the sound, but the more stuff I lost, the clearer it sounded. I rarely use more than a smattering of very short reverb. I tend to boost high EQ or add a little Sonic Maximizer for things that need more clarity (often, acoustic guitars, vocals and drum overheads) and cut low EQ for things that aren't bass drums and guitars to let them shine through. I also use panning more than most people here, because it gives the mix more room to breathe. Acoustic guitar is always done in stereo to give a bigger sound and clear out the middle of the sound spectrum for vocals. I use compression mostly on vocals and a little on guitar sometimes. Otherwise, I just get out of the way and let the mix happen on its own. *shrugs*

It sounds like the producers I admire most on here (Dylan, Puce, State Shirt, etc) do the same for the most part. It's "Producer Zen": the less you try to control, the better off you are.

Thanks, BTW!

[erik]

Someone teach me the magic words!

As a member of the MTCA (Muffled Thin and Clumsy Association) I wouldn't mind learning some too. I would ask that people be as fuckall specific as possible (like however specific you THINK you are being, be MORE specific than that) and avoid saying stupid things like "Just experiment!" because experimenting reeeeaaaallly doesn't work for some people.

[ken]

I haven't posted this for a while, but anyone interested in learning how to EQ things might want to start here. I find it extrememly helpful.

http://www.recordingeq.com/EQ/req0900/primer.htm

In general, try to cut more than you boost. Take things away from one element and add them to another.

Learning compression is (I think) much harder. I tend to use a lot of presets for my mixes. I think the classic compressor from Kjaerhaus has good presets for vocals and acoutic gutiars. I tend to use PSP Audio's Vintage Warmer on the lower instruments, and they have excellent presets as well. If you get VW, I'll tell you my secrets.

http://www.kjaerhusaudio.com/classic-compressor.php

The thing about reverb and delay, especially on vocals, is that it should be felt, and not heard. You want to find that point in the mix where it almost disappears, and set your level there. Unless of course, you are going for an effected sound.

Be well,
Ken

[starfinger]

I got a lot out of this book:

http://www.amazon.com/exec/obidos/tg/de ... s&n=507846

It has a lot of filler.. stuff about mixing in surround sound and interviews with famous mixers, but the heart of the book i s a great description of the fundamental mixing effects and how to go about putting a mix together.

-craig

[Phil. Redmon.]

Ken-

That EQ primer is rad. Lots of handy reference in one place. You are an hero.

[Bell Green]

A friend of mine went from using a four-track to using hard-disk recording and software. At first he started fiddling with all these new parameters and found that he started straying from his music. The software kind of got in the way. Once he had learned to make good recordings ie decent signal/noise ratio, mic positioning, which he already had, but hard disk had a slightly different character, then he could go back to the music again. He also started messing around with loops, which he said was another distraction. At one point he almost thought about actually going back to the four-track as it was so much simpler.

In the end he used his computer recording set up just like he did the four-track. So no recording four bars and looping them or staring at audio editing windows, but getting back to where he had left off. So his computer became his new four track and the music was with him again. He rehearsed everything before takes and did overdubs the old fashioned way. He did things exactly as he did on the four-track and found their equivalents in the software. Just volume, pan, hi/lo eq. He used one new thing on his virtual channel strip on each new song. That way he learned a few new tricks, but built on the old.

I made the same shift myself a few years ago and came to that same point where the technology or my lack of understanding of the technology got in the way of me making music. I got lots of tips from different friends and read recording magazines. But in the end went to college to study sound engineering and music technology. That broke the deadlock. The barrier between me and the music was gone, and I could start writing again. Armed with greater knowlegde and skills, I could now, at last, have an idea in my head and know how to get there. Knowing who Nyquist was isn't going to make you a better musician or songwriter, but the machine will seem more like a friend than a monster.

You have interesting times ahead. One step at a time . . .

[Mostess]

The software kind of got in the way.

Thanks for the story. I know what you mean. But 4-tracking was no picnic, either. Every second I spent rewinding was a second I could have spent playing. And that was a lot of seconds! That stupid little box sucked plenty of soul out of my music.

I'm not going back to old cassettes in cracked cases slowly print-through-ing my musical history into hissy garbage. No more ghosts of previous takes haunting my final ones. No more wow and flutter. No more trying to replicate my favorite mix-down.

Thanks for all the help so far. This information is great. My big issue is where to start. Is it best to make the bass instruments sound good and then add the higher-frequency ones? Or make each instrument sound good alone, and then add them together. Is it useful to pan everything before setting the levels, or vice versa? When you change the EQ of one thing, it affects how everything else sounds; are there any good heuristics for knowing what to work on first?

[starfinger]

My big issue is where to start.

I'd start with the tracks that are most important to the song. I like to get the vocals sounding good, fit the rhythm section around those and then layer in the other parts.

It depends on what you want the listener to get out of it.

-craig

[Bell Green]

Where to start?

Guess you have left the old four-track behind then. Ok, a few top tips then. Some of these are mine*, and some from my tutor** who has mixed many top bands.

*Leave a gap between recording and mixing. At least a day if you can, but not always possible with songfight. DON'T record and mix at the same time or mix as you go. Bad habit to get into, but can save time.

**Get the best recordings that you can in the first place. Mixing is about problem solving. Apply signal processing sparingly or you will end up in processing hell.

**Listen. Don't do anything except listen. Then decide where you want to go with it. There are two types of mix - open or closed. Everything else is a variation on this. It's tempting to start touching faders and knobs when listening, but just listen. You'll start to get a clearer and clearer picture (sound I suppose) of where the mix will end up. Keep doing this until that idea is firm. If the idea isn't firm then you will lose yourself. Once you start moving faders the image can get obscured. He said that this was pretty much the art of mixing - being able to hold that idea or image in your head, while you start moving faders etc.

*How to listen. Set up your reference monitors in an equilateral triangle with your head. So your speakers should be the same height as your ears. The distance between each speaker and your head should be same. Cover the screen if you need, don't look, listen. You need fresh ears, so that's partly why you need the gap. Close your eyes if you need and listen as many times as you need. It's same way that an interior designer comes into a room and can see what it will look like when they are done. Open mix, or closed mix? That's the first criteria when listening. Make notes - drums too quiet, vocals too loud in the chorus. If you don't get it, then take a break, go for a walk, come back to it. Don't touch, keep listening.

*The basic balance. Once you have an idea of where to go with the song and it's firmly in your mind, get a basic balance. Only use the faders and get everything sounding as best as you can with just that. Try different overall volumes (Fletcher-Munson curves).

Ok, more later. Time for dinner.

[Bell Green]

*continuation of the basic balance. If your idea is going to be an open mix, then you want your "band" to sound as though they are in a big room or hall. In a closed mix, in a small jamming room or club. The idea is to get it to sound "real". So the basic balance is to get an average balance of volumes so that all of the parts can be heard clearly, without resorting to any special effects or signal processing. Do the best you can with just the faders. Ok, it changes through the song, but don't worry about that just yet.

*less is more. Better to cut than to boost. Compression and Eq. In a nutshell one is to do with loudness and the latter with the timbre or frequency spectrum of the sound. To get the basic balance even better and clearer notice what you can hear clearly and what you can't. You can make the "treble" and "bass" instruments more treble and bass. Eg roll of the top end of a bass guitar (turn down the treble) and roll off the bottom end of a rhythm guitar(turn down the bass). But where? Which part of the spectrum? There are no hard or fast rules. The key of the song can give you a clue. Each note has a frequency. What is the highest note that the bass plays? What is that in cycles per second? So cut anything above that. In that way you can use eq musically. Compression is too big a subject to go into here. Just saying that it makes things louder will only confuse you. It makes things sound fat???? Use sparingly. If you can hear it, then it's too much.

*Automation. This is the bit that you couldn't do on the four-track, or at least you had to do it live. The changes that occur throughout the song. So far you have created a basic or average balance, but it needs to change. So you are trying to maintain that basic balance throughout the song, but then in the chorus that vocal could be 2dB louder and the rhythm guitar 3dB quieter. Well now that you have software, you can record your fader movements. Go into latch mode and record away. Push up that vocal fader at the start of the chorus and bring it down at the end. Switch to "read" mode and when you play back the song you'll see the fader move. In the arrange window you can graphically edit all your fader movements or create more.

**Automating the cuts. That's what my tutor used to call it. Basically it's about muting the tracks that aren't playing. Again the less is more philosophy of mixing. Let's say you have recorded a vocal harmony part which comes in at the end of each lead line. Well if there is nothing going on before and after then there will be noise - hiss. So mute that track just before that harmony vocal comes in and after it has finished. Do this with all the other parts. Record all the muting and unmuting. You'll see this muting and unmuting occur as you play the song. Your whole mix will start to sound much clearer now.

*Panning. Where is the hi-hat in a drumkit? On the left. But as you look at a drummer where is it? On your right. How far to the right? So pan that hi-hat to the right as far as it sounds correct. Where is the singer standing? You get the picture. Don't lose that initial image.

*Effects. The very last thing on the list. Use the effect sends rather than insert points which are reserved for compression and EQ. Set up effects busses and send your signal to that buss. The main effect is reverb. The point of it is to recreate that natural sound that you don't get in a dead room. With panning you dealt with left and right, with reverb its more a question of back and front. Again use just enough to create that "real" sound but no more. Which kind of reverb? Well, depends on that image that you have been holding in your mind - still have it don't you. Try and keep the reverbs consistent on all the instruments and parts. These can be automated as well, like an echo at the end of a vocal line.

*more listening. Take a break and then listen in different ways. Low, medium and high volumes. Listen from the next room. Listen through headphones. Listen in the car (a classic mixing technique). Listen through crappy speakers, if you have them.

Well that should give you a fairly good overview of the process. Listening, planning, basic balance, automation, effects and then more listening. That's my take anyways. Every engineer has their own process. Now zero the desk, take all the cables out of the patchbay and start again . . . only joking. If you are happy with it then you move to the next stage - mastering. The esoteric art.

They say that you shouldn't mix your own work and you should definitely not master it yourself. But such is the plight of the songfighter. Hope that helps.

[Caravan Ray]

Wow - where has this thread been all my life!

One question: If I record vocals and I want to compress, equalise and add reverb - is there any particular oder I should do these precesses in?

[joshw]

Wow - where has this thread been all my life!

One question: If I record vocals and I want to compress, equalise and add reverb - is there any particular oder I should do these precesses in?

I would EQ > Compress > Reverb. Definitely compress before reverb unless you're looking to use it more as an effect.

[jack]

josh brought up a real good point about panning, and using it to create more room for the mix to gel. for me, i always try to get optimal signal gain, as i can balance it out later but if i don't get optimal signal when i record, i re-do it. adding digital gain is crap if you can avoid it. i almost always fully pan drums and harmony vox slightly less than full, usually try to have stereo guitar tracks that i pan to about 48 each side to keep the vox and bass in the middle.

the other thing i like to do is before i make a final mix, i save a copy and bounce it. then i drop all my level sliders down and remix it all from scratch bringing up one channel at a time. create what i deem to be an optimal mix, bounce it and compare to the one i previously saved. the best one gets mastered.

[Sober]

eq primer = awesome. I'm gonna spend a lot of time funking around that site.

[deshead]

How about some love for a great old thread?

2 of my favorite mix tips:

1) If you don't have decent monitors, here are some good ways to check your mix:
- Crank the volume, then go to the next room over and close the door.
- Run the mix through headphones, but put the phones on the floor not your head.
- Listen in the car. Nothing emphasizes mud like 6-way speakers in an enclosed space.

2) Slowly turn the volume way down on a commercial mix. Notice that the first thing to disappear is the bass, and as the volume approaches 0, the last things you hear are the snare drum and the vocals. Make sure your own mixes do the same thing.

[jack]

How about some love for a great old thread?

2) Slowly turn the volume way down on a commercial mix. Notice that the first thing to disappear is the bass, and as the volume approaches 0, the last things you hear are the snare drum and the vocals. Make sure your own mixes do the same thing.

hey, this IS a great tip. thanks des!

[jack]

also, blue told me once he always mixes his drums full panned and high in the mix. i've followed that advice ever since. personally, i always fully pan the drums and center the bass. i also fully pan keyboards as a stereo track. i always try to use panning as a means of filling out the space so as many different instruments can be heard in the mix.

[boltoph]

a commercial mix....Make sure your own mixes do the same thing.

Yeahman. Even more so, do this during mastering. I listen to my favorite commercial tunes and compare mine to them, A/B style back and forth. This should reveal EQ differences between your mix and the pro mix. EQ plays a big part in mastering. A lot of mixes can be made real loud despite using any compression, if the instruments all have a combined eq that's nicely balanced across the entire frequency range. (and don't contain ecclectic spikes in sound, etc. ) If I don't do the comparison, my bass-loving ears tend to make everything way too bassy.
Use a frequency analyzer, if you wanna be real organic and such, manually lower the volume on any spikes in your wave editor, and give it a final limit but not much.
I've spent way too much time overdoing compression /limiting when I should've been focusing on the EQ of the mix.
Des you used to have a link on your site that gave a great chart that would tell you what frequencies are good for what instruments, but i couldn't find it now...hmm

Anyway i forgot to add, don't listen to me. I don't know what I'm doing...the last projedc I was involved on here is pretty much the roughest, quietest mix that will ever be made public by me...that one doesn't count .

[boltoph]

speaking of what we were just talking about, I'd say if you feel you must do a mix early on in the recording cycle (which is pretty much always the case for me, I mean just listening to your last take inspires a hundred mixing ideas), try to just focus on your tracks for this session, then finish your recording for the night, take a break, and then do an early mix. At least for me I don't even usually start vocals (except for ideas, sometimes a raw acoustic guitar / vocal track) until later on in the recording phase. So it's always good to hear a mix over a day or two to get your vocal ideas straight. But separation of recording tracks / mixing definitely needs to occur. I think it helps the rockin'

[Lunkhead]

Dang, the EQ primer Ken linked to many moons ago is now only available to paid subscribers. That sucks!!

[jb]

Dang, the EQ primer Ken linked to many moons ago is now only available to paid subscribers. That sucks!!

http://www.recordingwebsite.com/articles/eqprimer.php